The role of frequency content of environmental variability on extinction risk of age-structured populations: Spring-run Chinook salmon as an example

Background/Question/Methods

The role of the “color” of environmental variability (e.g., “white” or “reddened”) on population extinction risk has been examined, but the specific influence of spectral sensitivity of age-structured populations on extinction risk has not. In particular, age-structured populations are more sensitive to environmental noise at time scales near their mean age of spawning and to long time scales. This “cohort resonant” effect increases when populations decline. Because cohort resonance shapes the spectrum of population variability, it could influence extinction risk. Here we examine how the spectrum of environmental variability influences population variability and extinction risk in an age-structured population using the life history of spring-run Chinook salmon as an example. We used an age-structured population model with a Beverton-Holt stock recruitment relationship to demonstrate how the spectrum of environmental variability influences variability of spawning female abundance and the probability of quasi-extinction. Population simulations were run using time varying survival rates with different environmental spectra to investigate the influence of low-frequencies, frequencies equal to the mean age of spawning (generational time scales), and “reddened” noise.

Results/Conclusions

Simulation results showed how spawning female abundance was sensitive to variability at both generational and long time scales. Model runs of spawning female abundance forced only with noise at generational time scales had a higher standard deviation, but a lower coefficient of variation (CV) and did not experience quasi-extinction. Low frequency forcing and greater redness of environmental variability increased both the CV of spawning female abundance and quasi-extinction risk. Adding noise at generational time scales and higher frequencies generally reduced quasi-extinction risk compared to populations forced with low frequency noise. The spectrum of environmental variability matters in understanding the variability and, therefore, the extinction risk of age-structured populations in terms of cohort resonance. Specifically, environmental variability on generational time scales does not appear sufficient to increase probability of quasi-extinction, even though it had a great effect on the variability of abundance. However, cohort resonance also involves elevated sensitivity to low frequency environmental variability, which seems to be necessary for substantial quasi-extinction.